Value of Information Methods for NHLBI Trials

 描述(由申请人提供)：NHLBI在确定如何最好地分配随机临床试验(RCT)的投资方面面临重大挑战。目前，没有一致的 评估和比较NHLBI建议的RCT的机制，以最好地确定将其有限的美元投资于哪里。决策科学中的新方法可以提供有价值的信息，以加强NHLBI的RCT优先处理过程。该项目的总体目标是开发一套全面的实用信息价值(VOI)方法工具包和相应的软件，供临床研究人员和资助者随时使用，以评估与心、肺和血液疾病相关的随机对照试验的价值。此外，我们将开发标准化表格，报告这些计算的每个组成部分，以提高结果的透明度和传播性。使用这些方法，人们可以估计一项完美研究的价值，这项研究将消除所有不确定性，并立即被翻译。这样的估计将为区域现金转移的价值提供一个上限。或者，人们可以用部分消除不确定性的拟议设计来估计随机对照试验的价值。在其他应用中，人们可以为建议的研究得出最佳样本量或比较替代设计(例如，自适应设计与传统设计)。这些来自不同RCT投资的预期风险和价值的信息和展示可以帮助NHLBI对其投资组合进行量化可视化，并随着时间的推移优化该投资组合。

 DESCRIPTION (provided by applicant): NHLBI faces major challenges in determining how best to allocate investment in randomized clinical trials (RCTs). Currently, there is no consistent mechanism to evaluate and compare proposed RCTs for NHLBI to best determine where to invest its limited dollars. Novel approaches in decision science can provide valuable information to enhance the RCT prioritization process at NHLBI. The overall objective of this project is to develop a comprehensive toolkit of pragmatic value of information (VOI) approaches and the corresponding software that can be readily used by clinical researchers and funders in order to assess the value of RCTs related to heart, lung and blood diseases. In addition, we will develop standardized tables that would report each component of these calculations for an RCT to improve transparency and communication of the results. Using these methods, one can estimate the value of a perfect study that would eliminate all uncertainty and be translated immediately. Such an estimate would provide an upper bound to the value of the RCT. Alternatively; one can estimate the value of an RCT with a proposed design that would partially eliminate uncertainty. In other applications, one can derive optimal sample sizes or compare alternative designs (e.g. adaptive versus traditional) for a proposed study. Such information and presentations of expected risk and value from different RCT investments can help NHLBI with a quantitative visualization of its investment portfolio and optimize that portfolio over time.